Introduction
Childhood due to fast growth which is accompanied by biologic, psychologic and emotional changes increases nutritional needs. Attention to the natural development of children through healthy diet and continuous health support presents effective investigating for the health in future [1, 2]. In recent years, lifestyle changes have caused different societies to face problems of overweight and obesity in all ages especially during childhood and adolescence [3]. Diet in spite of its relation to growth and development may cause some health related problems during childhood [4]. Proper dietary habits in childhood may prevent diseases such as cardio-vascular [5] and osteoporosis in adulthood [6]. Improper dietary habits especially snacks with high fat is one of the most common factor of overweight and obesity in childhood. Study results in Rasht city indicated that the prevalence of overweight and obesity in the students of primary schools was 26.5 and 10.5 percent respectively [7]. Another study in Tehran showed significant statistical differences in frequency of snacks consumption in the morning, evening and free times and overweight and obesity [8]. One of the dietary problem of children is eating unhealthy snacks [2]. Results of a survey in Tehran [9] and another in Ghazvin [2].indicated that most students use sweets for snack. Education affects on the attitudes’ students significantly [2]. The main feature of unhealthy snacks is its low dietary value and high calorie with salt or sugar [2]. It is very important to correct nutritional status during childhood because it will become hard during adolescence since they do not comply with dietary change recommendations [4].  
Educational intervention can cause significant changes in students’ attitude [2]. Unfortunately most educator-led school based intervention has been ineffective at reducing health problems [10]. One of the effective ways of attaining educational goals is benefitting from children as teachers [1]. Educated and trained peers can communicate with the same age group in a way that health personnel, teachers and parents can not [2]. The child to child teaching approach was first introduced by David Morely in 1978 [10]. which emphasizes on child to child approach as one of the components of health promotion strategy [11]. School health based strategy started as an approach since 1990 and strongly supported by WHO ever since [11]. Study results conducted in Sanandaj city with child to child approach on dental health promotion indicated that peer-education group was more effective than trainer education group in performance domain [10].
A child to child approach study in relation to health of elementary students in Canada showed no significant difference in younger students after intervention but it was significant in older children [12]. Other studies also found the effect of teaching in peer groups [13, 14]. But in peer group teaching on Tuberclosis in adolescents, educator to child teaching was more effective [15]. In attention to controversial results regarding the effect of child to child teaching, the present study aimed to determine the effect of child to child teaching approach in comparison to health educator to child teaching approach on nutritional status in elementary school students. The study results may help authorities to plan to benefit from childrens’ participation in teaching, especially nutritional status in schools and take a step for children’s health promotion.
Materials and Methods
This is an educational trial study samples included 452 nine to eleven year old girls and boys from third, fourth and fifth grades from private and public schools of districts 1 and 2 in Rasht city (3 schools from each district) were chosen by random-cluster sampling method. Students’ food frequency condition was assessed by observational checklist and continued until sample size (204 students) with unfavorable food frequency achieved.  Then the 204 students were divided into three child to child, health educators to child and control groups. The first two groups received teaching intervention and control group received no intervention. The entrance criteria included enrollment as student in 3rd, 4th and 5th elementary schools and lack of disease and those not willing to participate in study were excluded. Based on unfavorable food frequency report in both study groups (P2=58.3% and P1=74.6%) according to Mohamadi study [2]. with 95% assurance and 5% devite, the sample size in each group was consisted of 68 students. 
Data-collection tool consisted of two sections: demographic questionnaire (sample age, sex, grade, and parents’ level of education, occupation, and number of children in family, number of rooms in house, child’s pocket money per week, and child’s view of his / her weight) and second section included observational checklist of food frequency. The number of primary sample size was used to determine the nutritional status before teaching intervention by completing 54 item of food frequency checklist with the supervision of researcher and teacher and three selected students from each class during school hours for one week. The checklist was completed by mothers at home for snacks taken during evening and holidays for one week. The observational checklist based on Food Frequency Questionnaire (FFQ) which its psychometry was done in Iran [16] and then completed checklists by mothers and researcher were collected and scored by researcher to determine favorable and unfavorable food frequency. In checklist, items of authorized and unauthorized snack-foods with a variety of four main groups of food pyramid were listed. Students in case of consuming variety authorized snack-foods obtained 0.5 point and if the diversity of snack-foods is not met or the students consume the unauthorized ones, obtained zero point. If samples use 28 authorized variety snacks or same valued nutritional snacks seven times or more during one week obtained satisfactory score of 14 and scores higher indicated favorable nutritional status and scores lower than 14 showed unfavorable nutritional status. 
According to the score of food frequency, students were divided into two groups. First group consisted of students with favorable food frequency that did not enter in the educational intervention and second group with unfavorable frequency were eligible for educational intervention. From primary sample chosen, 157 had favorable frequency and 295 had unfavorable frequency from which 204 students from selected schools were divided into three equal groups of 68 students in child-to-child teaching, educator to child and control group. Teaching intervention was conducted for educator to child and child-to-child group while control group received no intervention. After intervention, because of uncompleted check list by mothers at home, 6 students in child to child group and 10 students in control group were excluded. And 62 students in child to child and 58 in control group and 68 in educator to child group remained in the study (Figure 1). 

Children from each grade were selected in attention to the highest score obtained on checklist and comments of school principles and finally six students with best rhetoric were chosen for child to child teaching. Teaching with coordination of school principals were conducted in mid hours for thirty minutes in three days in one of the classrooms of boys and girls schools separately using pamphlets, pictures, and posters. Training sessions included introducing food pyramid, importance of snacks and authorized and unauthorized snack-foods, micronutrient and it role in snacks and fruit festival (each student brought one or two fruit and shared it after playing and praying). Teachings were presented by educator to child and child to child (each student taught his level grade) separately for three sessions each lasting 30 minutes for three grade levels (Table 1). Time table of program included two weeks for the first step (finding problem), two weeks education, and three weeks for follow up and two weeks for gathering checklist after teaching intervention lasting from March to May in 2015. After completing teaching intervention and last step observation checklist, the teaching booklets were given to the samples in control group. 

After three weeks, the observation checklist was completed again by researcher in school and mothers at home and the score of food frequency computed and determined the favorable and unfavorable cut-off points. Before intervention, the normality was surveyed by Kolmogorov and Smirnov test and showed that nutrition score does not obey normal distribution and data obtained were analyzed by SPSS 20 and Chi-Square, Mann whiteny and Kruskal Wallis tests. This research rejestered by IRCT N1 138706091126. Figure 1 runs in three groups of players to child, child to child, and control the nutritional status of primary school children.
Results
Findings indicated no significant difference between students’ nutritional status and variables such as type of school (public or private), sex, father’s job, parents’ education, child’s pocket money, number of rooms in the house, number of persons in the house in all three stydy groups. Study results showed that the nutritional status of 295 children (65.3%) was unfavorable. Findings showed significant difference between the three grade levels and students’ nutritional status (P=0.007). In fact the favorable frequency of third grade was more than other grades. Children’s nutritional frequency status before teaching intervention, based on mean score, was less than 14 and 100% of samples had unfavorable food frequency (Table 2). 

Based on results, even though the comparison of mean score difference of nutritional status in all three groups was not significant (P=0.089) but the changes in mean score of nutritional status was significant (P=0.0001). The greatest change was in child to child teaching group and the least change was in control group (Table 3). 

The mean score changes of nutritional status before and after teaching in child to child group in relation to educator to child group was significant (P=0.001) as child to child teaching group showed more teaching changes (1.19±0.37). The comparison of score changes in child to child and control group was statistically significant (P=0.0001) as child to child teaching group showed more teaching changes (1.98±0.39). The comparison of score changes in educator to child and control group was not statistically significant (P=0.179) and in comparison of three study groups as regard to changes in mean score of nutritional status before and after teaching, the child to child group was better than the other two groups. 
Table 4 shows that from 62 study cases taught by child to child approach, 33.9% had favorable nutritional status after teaching, from 68 students in educator to child teaching group, 36.8% had satisfactory frequency and from 58 students in control group who received no teaching intervention, only 14.3 % (9 students) had favorable nutritional status. Based on results, score of students’ favorable food frequency increased in all three groups and educator to child group was better than other two groups and it was statistically significant (P=0.014). While comparing the three groups regarding favorable frequency, the educator to child group was more successful (Table 5).  

Discussion 
Results indicates significant effect of teaching program on nutritional status of students in intervention group as compared to control group. In addition, most study cases had unfavorable nutritional status. This study finding is similar to some other studies on students’ food frequency status in Iran [2, 17, 18]. Previous and present study results showed students’ food frequency has unsatisfactory trend. In researcher viewpoint, students are using more low nutritional value snacks and parents are spending less time to prepare quality snacks due to their busy life. 
Results showed statistical significant difference between the three school grades as regard to students’ nutritional status, in fact the third grade students had more favorable food frequency than other grades. There was no significant difference between students’ nutritional status and variables such as type of school, sex, father’s job, and parents’ education, amount of child’s pocket money, number of rooms in the house and number of persons in the house. 
Results of some studies indicate that with increase in BMI, child and adolescents’ diet regime changes [19, 20]. Study results indicates more favorable food frequency in lower grades and unfavorable frequency in upper grades. It seems that after increase in age and BMI, changes in dietary regime is observed and as one nears puberty, food frequency gets more unfavorable.
There was significant difference in mean score changes of nutritional status before and after teaching in all three groups and least change was reported in control group that received no intervention. Study findings of Najar Lashkari on effect of child to child teaching health knowledge domain showed that the mean score of knowledge in intervention group of students increased [1]. In a study entitled “effect of peer teaching based on Health Belief Model on nutritional behavior of male elementary students” indicated that peer teaching significantly improved nutritional behavior of fourth grade students [13]. In attention to the results of various studies based on child to child teaching approach, it is necessary to assess the effect of peer practice in different health fields in schools and therefore benefit from this vast potential in teaching with less cost in schools.  
In comparison of child to child, educator to child and control groups, the mean nutrition score changes after intervention was statistically significant in child to child group and they had more educational changes than other two groups. Study conducted by Moeini showed significant mean change before and after intervention and in child to child group as compared to control group [10]. Baghian Moghadam study showed that peer teaching was more successful than other methods studied [14]. Other study reported the same results [21] but the result of Liu’s study [15] had controversial results. 
Of course as mentioned before present study found that the effect of educator to child teaching was more successful than child to child teaching on favorable nutritional status. It seems school health educators as trained adults have good effects on school health programs and their presence in school is justified. And it appears that assessing the effect of peer teaching as compared to school health educators needs more studies. In attention that students’ family did not receive any teachings and this matter could effect child’s snack nutrition status and also parents’ emotional condition while completing observation checklist at home could effect the results which was out of researcher’s control. Therefore it is suggested to compare the effect of family based teaching approach and peer teaching approach on students’ nutritional status. 
In conclusion, this study indicated that in achieving favorable food frequency, which was the main goal of the research, the educator group was more effective than peer group. But changes of nutritional behaviors in peer group was more successful than educator group. Our results seem to confirm previous research findings of peer- led influences on health education. Limitation of the study, we need to evaluate in long length time after two month that summer holiday started and we could not evaluate nutritional behaviors’ students in long length time.
Acknowledgments
This article is extracted from a MSc. thesis in Guilan university of Medical Sciences with 92237 ethics number, which is founded by Social Determinants of Health Resarch Center (SDHRC). The authors are grateful to all participants and teachers and managers and research and technology affairs Guilan University of Medical Sciences.  
Conflict of Interest
No conflict of interest has been declared by the authors. All authors have agreed on the final version and meet at least one of the ICMJE authorship criteria, including substantial contributions to conception and design, acquisition of data, or analysis and interpretation of data, drafting the article or revising it critically for important intellectual content.


References